The goal of this project is to better define the specific nature of memory loss in patients with Amnestic Mild Cognitive Impairment (a-MCI), a group enriched in patients with early Alzheimer's disease (AD) pathology. This work will aid in discrimination of memory loss due to AD pathology versus that of aging. Dual-process models of memory argue that recognition is subserved by recollection and familiarity. Both are impaired with early AD while only recollection is affected by aging. Little is known about the integrity of these processes in a-MCI. Therefore, recollection and familiarity will be measured by behavioral means in healthy elderly subjects and patients with a-MCI and mild AD. These estimates will then be related to (1) event-related potential (ERP) correlates of these processes and (2) medial temporal and cortical volumes using quantitative MRI techniques. As the earliest pathology of AD involves medial temporal structures, it is hypothesized that behavioral, ERP, and MRI measures will reflect involvement of this region in a-MCI and predict conversion to AD in longitudinal assessment. Specifically, familiarity, a form of memory thought dependent on perirhinal cortex, is expected to be impaired in a-MCI relative to elderly controls. Further, ERP correlates of recollection will differ between healthy aging and a-MCI/AD, reflecting prefrontal dysfunction in the former and hippocampal/entorhinal pathology in the latter. Relation of behavioral estimates of recollection and familiarity with volumetric MRI and ERP will provide further understanding of the anatomic and electrophysiologic bases of these processes and the impact of early AD pathology on them. The training plan for this Career Development Award application will build upon a foundation produced by the candidate's prior work in the study of memory in AD using behavioral and ERP techniques. In addition to enhancing these skills, the candidate will learn a new modality of inquiry, quantitative structural MRI. Training in such a multi-modality approach will allow for the candidate to address scientific questions in a novel and flexible manner as he becomes an independent investigator. Alzheimer's disease is already a tremendous public health burden which is expected to grow dramatically in the next 30 to 40 years. Patients with minimal impairment may benefit the most from potential disease modifying interventions. Thus, the pursuit of tools for early diagnosis, as proposed here, is critical.